Project Summary Crossing over between chromosome homologs is critical for their segregation in Meiosis I. In baker's yeast most meiotic crossovers (COs) are formed in an interference-dependent pathway in which double Holliday junctions (dHJs), stabilized by Msh4-Msh5, are resolved through the actions of STR helicase/topoisomerase, Exo1 nuclease, and the DNA mismatch repair MutL? (Mlh1-Mlh3) endonuclease. Our work is focused on creating a model to explain how MutL?, in a mechanism that is currently unknown, acts to resolve dHJs. This work will help us understand the molecular defects that underlie human infertility and hereditary forms of colorectal cancer. Our studies indicate that MutL? endonuclease is activated in DNA mismatch repair (MMR) and crossing over through the formation of a MutL? filament. We are testing this model in three Aims. In Aim 1 we will employ electron microscopy and DNA binding assays to analyze yeast MutL? polymerization properties. Endonuclease assays will be performed to measure MutL? activity on DNA substrates containing mismatch loops and structures predicted to be recombination intermediates such as HJs and branched DNA structures. These studies will be performed with MutL? alone and in the presence of interacting factors. Residues required for MutL? interactions with these substrates and endonuclease cleavage sites will be mapped. We will also perform single-molecule analyses to determine how MutL? binds to DNA substrates such as loop mismatches and the role that the Mlh3 linker domain in MutL? plays in this process. In Aim 2 we will identify meiotic factors that interact with yeast MutL?. Using a structure-function approach we identified five CO proficient, MMR defective alleles and one CO defective, MMR proficient, mlh3 allele that does not disrupt MutL? endonuclease activity. Our 76 allele set provides a foundation for experiments that examine interactions between MutL? and STR, Exo1, Msh4-Msh5, Zip3 and newly identified components. We will examine the biochemical functions of Mlh1-mlh3 complexes containing mlh3 null and separation of function mutations. Mutant complexes defective in specific biochemical activities (DNA binding, ATPase, endonuclease) will be used as baits in genetic suppression (high copy, reduced dosage) approaches involving a large set of meiotic factors. Meiotic extract pull-down assays using wild-type and specific Mlh1-mlh3 complexes will be performed in parallel. Finally, we will perform epistasis analyses between mlh3 mutants and meiotic components to create a MutL? network to be tested in biochemical assays (Aim 1). In Aim 3 we will study why mouse MutL? has a linker domain seven times larger than the yeast domain. Linker domains in Mlh proteins undergo conformational changes and mediate DNA binding; interestingly genetic and cell biology studies suggest that the mouse linker also contains novel activities. We will purify mouse MutL? and linker domain deletion mutants and will analyze them in assays developed for yeast MutL? in Aim 1. This work involves a collaboration with Paula Cohen, who will perform genetic and cell biological analyses of mouse MutL?.